The Professional Body for Medical Genetics Recommends Against MTHFR Testing. Here's Why That Matters

MTHFR variants are among the most-tested and most-misunderstood in consumer genetics. The American College of Medical Genetics recommends against testing them in most clinical settings. What the gene actually does, and how to read a result honestly.

Someone is exhausted, has been for years, and has been told by a wellness site that a gene called MTHFR explains it. They buy the methylated folate. Sometimes they feel better for a while.

MTHFR may be the most over-interpreted gene in consumer genetics. It has been blamed for miscarriage, clotting, depression, autism, fatigue and chronic illness, and it anchors a substantial supplement industry.

So this post starts with the least commercial sentence available.

In 2013 the American College of Medical Genetics and Genomics issued a practice guideline titled, in full, "lack of evidence for MTHFR polymorphism testing" (Hickey et al., Genetics in Medicine, 2013, PMID: 23288205). Its recommendation was that MTHFR polymorphism testing should not be ordered as part of the clinical evaluation for thrombophilia, recurrent pregnancy loss, or in at-risk relatives. The position was reaffirmed in a 2020 addendum (doi:10.1038/s41436-020-0843-0, PMID: 32533132).

That guidance is from the profession that would benefit most from ordering the test. It is worth taking seriously.

What the gene actually does

MTHFR encodes methylenetetrahydrofolate reductase, an enzyme in the folate cycle. It converts one form of folate into the form used to convert homocysteine into methionine.

The best-known variant, C677T (rs1801133), reduces enzyme activity. People homozygous for the T allele have measurably lower enzyme function and, on average, modestly higher homocysteine. A second variant, A1298C (rs1801131), is frequently reported alongside it.

None of that is in dispute. The biochemistry is real and well characterised.

Why the recommendation went the other way

Three reasons, all of which generalise well beyond this gene.

The variant is extremely common. Depending on population, roughly 10% of people are homozygous for C677T. A "risk variant" carried by a tenth of humanity is not a rare explanatory finding.

The downstream effect is modest and manageable. Homocysteine elevation associated with the variant is small in folate-replete populations, and mandatory folic acid fortification has reduced it further in countries that adopted it.

Lowering homocysteine has not delivered the expected benefit. This is the part that settled the question. Trials lowering homocysteine with B vitamins have generally not produced the reductions in cardiovascular events or thrombosis that the hypothesis predicted.

A biochemical mechanism can be entirely real and still fail to translate into a clinical intervention that helps. MTHFR is the textbook case.

Where testing does have a place

Homocysteine can be measured directly, which is more informative than inferring it from genotype. Severe MTHFR deficiency — a rare condition, distinct from the common polymorphisms — is a genuine clinical entity. And folate status matters in pregnancy regardless of genotype, which is why folic acid supplementation is recommended broadly rather than by genetic testing.

If you are worried about clotting, recurrent miscarriage or cardiovascular risk, the evaluation belongs with a doctor and does not begin with MTHFR genotyping. This article is not medical advice and nothing here should change a supplement or a medication.

What the app reports

MTHFR Insights reports your genotype at the common MTHFR variants and explains what the research literature says about them. Like every app on the platform, it tells you where you would have scored had you participated in the original studies - it is not a statement that your own risk of anything is raised or lowered.

The developer's own disclaimer is unusually direct, and worth quoting: the application "is not a diagnosis, a prediction, or a predisposition score" and "is an implementation of the findings in referenced papers for research purposes only."

Read it that way and it is genuinely interesting: you are looking at a well-studied enzyme variant in your own genome. Read it as an explanation for how you feel, and you will be joining a very large group of people who were sold that story and did not get better.